INDEX

Explanations

phrases related to upper and lower levels or classes, and their associated qualities

oai_token-act-pair · gpt-4o-mini Triggered by @bot

New Auto-Interp

Top Features by Cosine Similarity

Comparing With LLAMA3.1-8B @ 30-llamascope-res-32k

Configuration

fnlp/Llama3_1-8B-Base-LXR-8x/Llama3_1-8B-Base-L30R-8x

Prompts (Dashboard)

24,576 prompts, 128 tokens each

Dataset (Dashboard)

cerebras/SlimPajama-627B

Features

32,768

Data Type

bfloat16

Hook Name

blocks.30.hook_resid_post

Hook Layer

Architecture

jumprelu

Context Size

1,024

Dataset

cerebras/SlimPajama-627B

Activation Function

relu

Embeds

IFrame

Link

Not in Any Lists

No Comments

Negative Logits

sse

-0.16

shed

-0.15

efe

-0.15

Ð´Ð°Ð»

-0.15

eland

-0.14

 Svens

-0.14

eph

-0.14

emu

-0.13

rgb

-0.13

eur

-0.13

POSITIVE LOGITS

most

0.44

MOST

0.27

-middle

0.26

 reaches

0.25

-most

0.25

cased

0.24

/l

0.24

class

0.24

ech

0.23

archy

0.23

Activations Density 0.031%

phrases related to upper and lower levels or classes, and their associated qualities

No Comments

No Known Activations